Originally appeared in: March 1999 issue, pgs 161-172

HYDROCARBON PROCESSING Reprinted with publisher’s permission.

Improve Vacuum Heater

Reliability
These design, operation and revamp
guidelines will help increase run lengths
and efficiency.
A. Garg, Furnaces Improvements, Sugar Land, Texas

To improve the reliability of vacuum heaters, a good heater

design is the first prerequisite. This must be followed by good heater
operation and maintenance. Together, they can go a long way towards
improving heater reliability and run lengths. The vacuum
heater is one of the most important heaters in any refinery. In
the last few years, vacuum units have become more important
due to heavier crudes being processed. The price differential in
crudes is also forcing refiners to switch to heavier crudes. This
trend is causing vacuum heaters to be overfired. This leads to
short run lengths and affects refinery profitability.
Many factors can improve vacuum heater reliabil-
ity. In most cases, improving the existing design and operation
can increase the run lengths of the vacuum heaters.
Run lengths. About 25 years ago, typical run
lengths of vacuum heaters were 18 to 24 months. Over time,
refiners have become more demanding on their units and the
standard slowly increased to 3 to 4 years and currently stands
at 5 to 6 years. Most refiners are trying to achieve these run
lengths in their vacuum heaters. A recent survey showed that
refiners are actully getting only 1 to 3 year run lengths from
their vacuum heaters. In two instances, the clients were
forced to shut down within a year of the heater's startup, and
had to decoke the heaters and start up again.
What is causing such short run lengths? One of the
major reasons for short run lengths is linked to coke formation
in the radiant section tubes. With excessive coke formation,
refiners cannot process the required furnace throughput and
have to prematurely shut down to clean the heaters.
Coke deposits. Heat is transferred from outside
the furnace tubes to the fluid inside by conduction through the
tube metal. Heating oil molecules to a high temperature causes
"cracking," with the subsequent formation of coke on the tube's
inner surface. The coke layer that is formed is a poor heat
conductor. It insulates the tubes and impedes heat transfer. This
causes the heater to be fired harder to maintain the process
fluid at the required outlet temperature.
Coke formation is an inevitable feature of vacuum
heaters. But, generally, it is an undesirable outcome of poor
design and operation. Coke can be laid down by flame impinge-
ment. It may be due to low flows on the tube side, which allows
a high metal temperature to be attained even without excessive
firing conditions. In either event, tube rupture is a foresee-
able outcome. Sometimes localized "hot spots" develop on tubes
where flame impingement or a partial loss of flow has occurred.
After initial lay down of coke in a heater tube, a
vicious circle has begun in which more and more coke is laid
down. Coking rate increases with the temperature in the tubes.
It keeps on depositing on the hotter layers of the existing per-
meable coke. Eventually, the coke hardens and becomes imper-
meable. The vacuum heater operation not only becomes ineffi-
cient (due to higher heat losses in the flue gases resulting
from the increased heat input), but also potentially danger-
ous. The danger lies in the excessively high temperatures that
the tubes may reach, which could cause rapid scaling of the
metal and a possible tube rupture.
It is extremely important to avoid flame impingement
and low flow through the tubes. In addition to an actual tube rupture,
tubes can sag or bow. A tube will sag under its own weight if it
becomes grossly overheated. Uneven coke laydown in a tube will
make one side of the tube expand more than the other side,
leading to bowing. Tube metals differ according to the type of
process and severity of duty, but they can all suffer damage due to
overheating. As a check against overheating, thermocouples are
usually installed to monitor the tube skin temperature at various
points within the heater. Infrared inspections can also detect hot
tubes. Good heater operations require continuous monitoring of
heater tube metal temperatures for an abnormal rise.
Two major factors are responsible for short run lengths of vacuum
heaters—poor design and poor operation.
Which heater is more reliable? This is a never-
ending debate between the horizontal tube and vertical tube
heater. Industry started with horizontal tube heaters and is
slowly moving towards vertical heaters. Vertical tube designs
have been tried in almost all services. Vacuum heaters are no
exception.
The typical horizontal tube vacuum heater is a single-
cell or twin-cell cabin heater. Radiant tubes are laid out hori-
zontally on both sidewalls. Fluid flow is either from top to bottom
or bottom to top. The burners are placed on the floor and are spaced
to provide uniform heat distribution (Fig. 1). Some older heater
designs have the burners firing horizontally on both end walls.
The heater's arch is either flat or sloped. Sometimes, radiant
tubes are also installed in the arch section. The convection sec-
tion is mounted on top of the radiant section. It consists of bare and
extended-surface tubes. Finned tubes are used in gas-fired heaters
and studded tubes are used in oil-fired heaters. The convection
section absorbs 20% to 30% of the total heat duty. Most modern vacuum

Furnace Improvements Services

www.heatflux.com
Improve Vacuum Heater Reliability
heaters have a provision for adding two extra rows in the convection
section, which come in very handy during debottlenecking. Flue gases
are cooled within 150°F to 250°F of the inlet fluid temperature.
Many vacuum heaters are provided with steam super-
heating coils in the convection section for super-heating column
steam. The stack is mounted on top of the convection section to
provide draft and dispose the flue gas safely. Fig. 1 provides a
cross-sectional view of a horizontal tube vacuum heater. One of
the distinct parameters in a vacuum heater is the operating condi-
tion at the vacuum heater outlet. The radiant coil is characterized
by the presence of increasing tube size towards the outlet. Typi-
cally, the last three or four tubes in each pass are gradually in-
creased in size to keep the fluid velocity below the sonic veloc-
ity.
The vertical-tube radiant heater is mostly a vertical
cylindrical firebox. The tubes line the sidewalls and the burners are
installed on the floor in a circle (Fig. 2). Since the tubes are installed
vertically, the plot area requirements are much lower (Fig. 3). The
vertical-tube design allows larger and fewer burners. The draft
available in vertical tube designs is mostly higher than box heaters.
The convection section is generally compact. Vertical-tube designs
are generally limited to a 60ft. tube length due to tube handling
limitations. The size of vertical heaters is limited to 150 MMBtu/
hr, although larger heaters have been built using a multicellular
concept. Vertical-tube heaters are compact, require less plot space,
and are more economical than horizontal-tube heaters. A number
of vacuum heaters have been designed and operated in the verti-
cal-tube configuration.
The major concern of process engineers with vertical tubes
is that two-phase flow is not well defined in the vertical tubes and
the upflow pattern is different than the downflow pattern. Two-phase
flow is more forgiving in horizontal tubes. Proper mass velocities
and coil design can take care of this concern. Vertical-tube design
is almost 20% to 30% cheaper than the horizontal tube design.
Operating either heater is equally demanding.
Requirements for good operations are the same in
both types of designs. Vertical-tube vacuum heaters require all
the burners to be operated for uniform heat transfer. In horizontal-
tube heaters, one of the burners can be shut down without affecting
the pass outlet tem-peratures. For any type of heater to have a
long run length, the heater design must be good. Vacuum heaters are
conservatively designed with low height/diameter (H/D) ratios. The
height of the radiant section is generally limited to 2.0 to 2.5 times the
radiant section width, or the diameter with vertical cylindrical
heaters.
Mass velocity. In both the horizontal and vertical
tubes, it is important to keep the tube side mass velocity as high as
possible. Higher mass velocities keep the tubes cooler by improving
the inside heat-transfer coefficient. Look at the two cases of tubes
(Table 1), one with 2
400-1b/sec-ft2 mass velocity and the other
with 200-1b/sec-ft velocity. One heater is a four-pass design and the
Furnace Improvements Services
www.heatflux.com 2
other heater has two passes.
Based on the data in Table 1, clearly the higher mass
velocity will always help reduce cracking and coking by
lowering film temperature.
Radiant heat flux. One of the important criteria in
vacuum heater design is radiant heat flux. It is defined as heat
transferred per unit of tube area. There are no fixed numbers for
vacuum heaters, but they are generally 8,000 to 10,000 Btu/hr-ft2.
Vacuum heaters are generally designed with lower heat flux as
compared to atmospheric heaters. This means that the fire-
box size will be larger, which translates into more uniform heat-
ing. Lower heat flux keeps the fluid temperature below the
fluid's cracking temperature. Table 2 compares three vacuum
heaters with three different fluxes. In dry vacuum heaters, heat
fluxes are even lower than the wet vacuum heaters.
Tube size. The number of passes determines coil
size in a vacuum heater. In a typical heater, 6-in. tubes are used.
However, vacuum heaters are generally designed with smaller
tubes, which provide more heat-transfer surface per unit vol-
ume, The inside heat-transfer coefficient is also higher in
smaller diameter tubes for the same velocity. However, smaller
tubes lead to more tube passes. And this increases the chance of
bad operations. A balance must be struck between tube passes
and tube diameters.
Outlet conditions. Typically, the column inlet
temperature ranges from 700°F to 750°F, and pressure ranges
from 2 to 4 inHg. Transfer lines are used to connect the vacuum
furnace outlets with the distillation tower. The vacuum heater
transfer line requires special considerations due to the fluid's
high specific volume at the tower conditions. Even with a low
fraction vaporized, dispersed flow conditions exist in the outlet
tubes and transfer line due to the low pressure level and high
vapor specific volume. If the transfer line is too small, a high
pressure drop will be created in the transfer line. Conse-
quently, the charge outlet temperature must be increased to
maintain the desired design flash-zone temperature. Higher
outlet temperatures lead to more furnace coking and crack-
ing. Pressure drop in the transfer line should be set so the re-
quired flash-zone conditions can be obtained without exces-
sively high furnace outlet temperatures.
Case study. In one vacuum heater, the client was experiencing
frequent coking. Analysis of the problem revealed that the flash-
zone temperature was 720°F and the vacuum heater outlet tem-
perature was 777°F. The client was experiencing more than 50°F
temperature drops across the transfer line. In this case, increasing the
Improve Vacuum Heater Reliability
transfer line size was the first step. The client ended up increasing the
transfer line from 12-in. to 24-in. This reduced the temperature drop
across the transfer line to 15°F.
Peaking temperature. Another phenomenon that oc-
curs in the vacuum heater outlet tubes is fluid temperature peak-
ing. This happens due to the pressure drop in the tubes and fit-
tings, which suppresses charge vaporization and results in a
fluid temperature increase. In most cases, the maximum bulk
temperature in the heater is generally higher than the outlet
temperature. Excess cracking increases the tendency for coke lay-
down. In a well-designed heater, peaking should not be more than
a few degrees.
Steam injection. Another distinct feature of a vacuum
heater is the steam injection in the heater coils. The presence of
steam in the heater helps reduce the hydrocarbon partial pressure.
It induces vaporization at lower temperatures, which reduces the
peak film temperatures. Steam injection in the vacuum heater re-
sults in extra charge vaporization. Injecting steam often means
that the pressure drop across the heater substantially increases.
Steam can be injected at the heater inlet, radiant inlet, the point
of vaporization or where the charge reaches its cracking tem-
perature. Steam injection is also beneficial in reducing tube cok-
ing in a limited way.
Design improvements. Several improvements can
help increase vacuum heater reliability. Most are based on reduc-
ing the heat load of the radiant section or eliminating flame impinge-
ment. The major methods are to:
Increase heat transfer area in the convection section
Install forced-draft burners
Add burners.
Increasing convection surface. It is easier to add
heat-transfer surface in the convection section than in the radiant
section. The convection section absorbs 20% to 30% of the total heat
duty. Increasing the heat transfer surface can reduce the flue gas
temperature leaving the convection section. The flue gas tem-
perature approach can be reduced to 150°F of the charge inlet
temperature. Increasing the convection section surface provides
two benefits. Higher efficiency will translate into lower firing
rates and enhanced heat-transfer surface will reduce the load on the
convection section. This can be done in the convection section
by:
1. Adding tubes in the convection section
2. Replacing bare tubes with extended-surface tubes
3. Substituting finned tubes for studded tubes
4. Change of service.
Two additional rows of tubes can be installed in the convection
section of most vacuum heaters without making a major
change. The inlet-piping terminals need to be relocated.
Case study. A vacuum heater was plagued with
short run lengths. An analysis of the heater revealed that it was
running at 15,000 Btu/hr-ft2. Flue gas leaving the convection section
was more than 1,100°F. High flue gas temperature was loading the
air preheater and induced draft (ID) fan. A high flue gas tempera-
ture to the air preheater was also leading to higher air preheat
temperature. This led to a further increase in the radiant heat
flux. The high rate of radiant heat transfer accelerated tube cok-
ing.
Furnace Improvements Services
www.heatflux.com 3
Solution. The convection section performance was
examined and it was recommended to maximize the number
of tubes in the convection section. This minimized the radiant
section heat load, kept tubes cool and improved the system's overall
operation and run length. The flue gas temperature exiting the
new convection was reduced to 800°F. The performance before
and after the modification is compared in Table 3.
Adding burners. One of the important requirements
in achieving a longer run length is a small flame size or elimina-
tion of any flame impingement. In many vacuum heaters, the ven-
dor supplies the minimum possible number of burners to keep
costs low. It always helps to increase the number of burners in a
vacuum heater to a practical maximum (Fig. 4). This increase
will improve heat-flux distribution, help uniformly fire furnaces
and improve run lengths. Smaller burners have shorter flame lengths
and the chances of flame impingement are eliminated.
Flow distribution. This is important in a multipass
heater. Flow maldistribution can cause coking in some passes.
The number of passes in a vacuum heater is determined con-
sidering the available pressure drop across the heater and outlet
velocity. The passes in a vacuum heater are generally an even
number to ensure uniform heat transfer.
Fluid being heated inside the tubes must be controlled
for efficient heat transfer and to minimize tube fouling and cok-
ing. Flow distribution at inlet is very important. All fluid passes
should have an equal amount of fluid passing through the tubes
(Table 4). In most liquid or fouling services, it is important to
have an individual pass flow controller to avoid imbalancing of the
flow due to coking or localized overheating (Table 5).
Many control schemes have been used to control process
flows to the heater passes. A common control scheme where the indi-
vidual pass outlet temperatures are controlled to ensure the
uniform outlet temperatures has been implemented in a number
of vacuum heaters. This scheme works fine as long as the ser-
vice is not fouling. With coking or fouling services, it is unsat-
isfactory because it tries to cut down on flow in the pass that is
coked, and the situation becomes even worse. The pass tends to
coke even more at the reduced flow.
Fluid flowing in the tubes should have an adequate pressure drop
in the fired heater to ensure good distribution in a multipass
Improve Vacuum Heater Reliability
heater. If the pressure drop across the heater is low, then there is a
chance for flow imbalance and a pass may run dry.
Firing controls. Three parameters that must be controlled and
monitored are:
Fuel gas/fuel oil pressure
Excess air
Draft, in the furnace.
The best way to improve vacuum heater run length is
to uniformly fire the heater. Also, keeping all the burners in
line helps. It is important to maintain the burner in good condi-
tions. The registers should be operating and not stuck. In a natural-
draft heater, keep all registers with the same opening. In forced-
draft firing, the burner registers should be fully open.
Another item required for good operation is a reliable
stack damper. It is important to have a good control of draft in the
heater. It cannot be achieved without a good stack damper.
Over firing. One common problem is over firing. This
leads to positive pressure at the arch and flame impingement. If the
furnace is facing positive pressure, then the stack needs to be
modified. The vacuum heater was facing positive pressure at the
arch. Providing an inde-pendent stack on vacuum heater modi-
fied it (Fig. 5).
Excess air control. This has always revolved around
reducing excess air. Operators have always reduced excess air
to the lowest possible level with scant regard to burner perform-
ance. Burner performance is vital to improving the heater reliabil-
ity. Excess air should be increased until the good burner performance
is achieved. This, in turn, improves the vacuum heater's run length.
Air preheating system. Adding an air preheater is one
Furnace Improvements Services
www.heatflux.com 4
way to improve capacity and efficiency. However, an increase
in radiant heat flux and a corresponding increase in tube metal
temperature is often overlooked. In most cases, the average heat
flux goes up by 8% to 10%. This increases the tube metal tem-
perature correspondingly. Any time an air preheater is added
to a vacuum heater, the complete system must be reviewed
and rerated to ensure reliable operation. Another problem is that,
when adding an air preheating system, two fans are introduced and
the control system becomes much more complex. This affects over-
all system reliability. Fans require continu-ous attention and
maintenance. It is always cheaper to buy reliable low-speed
fans for long run lengths.
Case study. A vacuum heater was fitted with an
air preheating system. The client was pushing the vacuum
heater hard and the excess oxygen was running close to 0% to
1%. The burner flames were long and lazy and the skin
temperatures were running high. An inspection revealed that
the heater was running short on air. The field survey deter-
mined that the air preheater bypass was closed. Opening the
air preheater bypass increased the excess air supply at the
cost of losing some flue gas heat. The excess air was in-
creased to 20% to 25%. This reduced the flame sizes drasti-
cally, eliminated the flame impingement and brought down
the tube skin temperatures quite a bit. This prevented a
shutdown and prolonged the run length.
BIBLIOGRAPHY
Garg, A., "Better burner specifications," Hydrocarbon Proc-
essing, August 1989.
Garg, A. "Better low NOx burners for your furnaces,"
Chemical Engineering Progress, January 1994.
Garg, A. "How to boost performance of fired heaters," Chemical
Engineering, November 1989. Garg, A. "Optimize fired heater
operations to save money," Hydrocarbon Processing, June
1997.
Garg, A. "Trimming NOx from furnaces," Chemical Engi-
neering, November 1992. Garg, A., and H. Ghosh,
"Every BTU counts," Chemical Engineering, October
1990. Garg, A., and H. Ghosh, "Good fired heater specifica-
tions pay off," Chemical Engineering, July 9, 1988.
Ashutosh Garg is currently a thermal
engineer at Furnace Improvements,
Sugar Land, Texas. He has more than 24
years of experience in design, engineering
and troubleshooting of fired heaters and
combustion systems for the refining and
petrochemical industries. He graduated
from Indian Institute of Technology, Kanpur,
India, in chemical engineering in 1974. He
started as a graduate engineer in an
ammonia plant. This was followed by six years at KTI
India and eight years at EIL, New Delhi, in the heater
group. He joined KTI Corporation, San Dimas, California, in
May 1990 and moved to KTI Houston in 1992. He has
published several papers on fired heaters and burners.
He is a registered professional engineer and a member
of AIChE and ASME. He is also a member of the API
subcommittee of heat transfer equipment and is on the
task force for the new API standard for flares.